Using Photographs of Contrasting Contextual Complexity to Support Classroom Transitions for Children With Autism Spectrum Disorders

Abstract

A single-subject, alternating treatment study compared the impact of two types of photograph displays of contrasting contextual complexity. The study examined the impact of high-context and no-context photographs, displayed on an iPad, on the ability of three preschool children with autism spectrum disorder (ASD) to transition to play activities. The influence of the two photograph types on the children’s duration of transition time, number of prompts required, and type of prompts were measured. Results indicated that both photograph types provided helpful support for the children’s transition time to play activities. Results for duration of transition time and number of prompts required during transitions suggest that the participant with more challenging needs performed more quickly and independently using high-context photographs. Results suggest the photographs that have contextual information matched to a child’s needs can provide salient information and support their transition to activities in familiar settings.

Keywords

autism iPad augmented input contextual photographs

The Influence of Photographs of Contrasting Contextual Complexity for Children With Autism Spectrum Disorders to Support Classroom Transitions

Autism spectrum disorder (ASD) is a group of developmental disabilities that had a recent definition revision in the Diagnostic and Statistic Manual of Mental Disorders (DSM-5; American Psychiatric Association, 2013). The definition now represents a single diagnostic category that reflects a continuum of symptoms. The spectrum of the disorder includes individuals with a range of abilities who share common characteristics in communication, social skills, and behavioral challenges. Individuals with ASD have communication deficits in both expressive and receptive communication skills and may benefit from augmentative and alternative communication (AAC) interventions.

Mirenda and Brown (2009) stated that all children with ASD have some type of communication impairment, and that the specific communication challenges experienced can vary from person to person. Individuals with ASD exhibit strengths in visual processing skills (Schopler, Mesibov, & Hearsey, 1995) and are strong visual learners (Mirenda & Brown, 2009). Given this well-known characteristic, it is logical that practitioners use visual approaches with individuals with ASD. Some individuals with ASD may benefit from AAC for expressive means and may also need augmented support to comprehend the language and information conveyed to them (Wood, Lasker, Siegel-Causey, Beukelman, & Ball, 1998).

One type of AAC intervention that may be beneficial to individuals with ASD can be provided with visual scene displays (VSDs). These displays depict people, actions, or objects associated with specific events or activities (Blackstone, 2004; Shane & Weiss-Kapp, 2008). Images (e.g., photographs, digital pictures) in VSDs can depict a generic context or a personalized context. A central unique aspect of VSDs is that elements depicted are within a coherent, integrated visual image (Beukelman & Mirenda, 2013). Single, isolated symbols commonly used, may have limited impact on comprehension for individuals with ASD. It may be beneficial for individuals with ASD to be provided with VSDs via photographs as contextually rich images that are also personally relevant.

AAC and Visual Supports

A characteristic of individuals with ASD is gestalt processing (Prizant, 1983). Weissling and Beukelman (2006) use the same term to describe visual scenes: “The presence of personally and contextually relevant information can trigger the gestalt processing of a scene” (p. 15). Although much is still unknown about the manner in which individuals with ASD process information, it is important to investigate an approach that has potential to suit their learning and processing styles. Visual support strategies appear to alleviate at least some of the confusion, anxiety, and frustration that many individuals with ASD experience when they encounter unexpected events (Mirenda & Brown, 2009).

VSDs provide individual contextual information relevant to a specific person. Whole scenes are likely to provide greater meaning of schematic or experiential content to persons with ASD and may benefit those who may not comprehend the semantic meaning behind images of isolated nouns (Shane, 2006). High-context photographs can be considered a scene display of personal relevance and may improve comprehension for children with ASD.

The purpose of this study was to examine the impact of high-context photograph with no-context photographs on the amount of time and prompts needed for a group of children with ASD to transition to interactive play activities with classroom peers.

Method

Participants were recruited through local public school Early Childhood Education Programs. Teachers who served children with disabilities between 3 and 5 years of age contacted the investigator (second author). Observing confidentiality, the investigator interviewed the teacher to verify that there was a student who had these characteristics: (a) a medical diagnosis or educational verification of autism, (b) needed support to move to classroom activities, and (c) had expressive, spontaneous communication that was less than 25 single spoken words or other symbols and less than 10 combined spoken words or other symbol combinations.

Informed parental consent was obtained for six children, and three met the specific criteria for inclusion in this study. For each of three participants with ASD, two peer participants were selected for the study through teacher nomination and informed parental consent was conducted. The peers without disabilities were selected because of their ability to follow teacher instructions, to share toys, and to wait, when asked.

Assessment Procedures for Participants With ASD

Assessments were conducted by the investigator to document the current level of adaptive functioning and communication development for each participant. The investigator completed the Vineland II Adaptive Behavior Scales–Second Edition (Sparrow, Cicchetti, & Balla, 2005) and the Communication Matrix (Rowland, 2004) through teacher interviews. A symbol assessment directly with participants to document their ability to match objects to photographs using a single-symbol-to-multiple-object matching format adapted from Mirenda and Locke (1989). A teach-test session was conducted if a participant did not reach 70% accuracy. A summary of complete assessment results is provided in Table 1.

Table 1.

Description of Participants With ASD.

		Vineland II			Communication matrix
Participant	CA	Domain	SS	Percentile	Function	Communicative competence level
Mario	3-3	Communication	65	1.0	Refuse	III
		Daily living skills	60	0.4	Obtain	III
		Socialization	60	0.4	Social	III
		Motor skills	78	7.0	Information	Below I
		Composite	61	0.5
Cody	4-0	Communication	67	1.0	Refuse	IV
		Daily living skills	68	2.0	Obtain	V
		Socialization	66	1.0	Social	III
		Motor skills	70	2.0	Information	IV
		Composite	64	1.0
Sage	4-5	Communication	44	<0.1	Refuse	III
		Daily living skills	43	<0.1	Obtain	III
		Socialization	48	<0.1	Social	II
		Motor skills	70	2.0	Information	Below I
		Composite	45	<0.1

Note. Vineland II Adaptive Behavior Scales (Sparrow, Cicchetti, & Balla, 2005). Communication Matrix (Rowland, 2004). Level I: Pre-Intentional Behavior; Level II: Intentional Behavior; Level III: Unconventional Communication; Level IV: Conventional Communication; Level V: Concrete Symbols; Level VI: Abstract Symbols; Level VII: Language. ASD = autism spectrum disorder; CA = chronological age; SS = standard score.

Description of Participant 1: Mario

Mario was a 3-year, 3-month-old Hispanic male with an educational verification of autism. He exhibited some self-stimulatory behaviors that included shaking his head side-to-side, gazing at his hands in the air, and body-posturing with one or both arms. He displayed emerging intentional communication behaviors and presymbolic communication (Communi-cation Matrix; Rowland, 2004). He primarily used unconventional communication during the study, using behaviors that included turning away or closing his eyes for protest, touching or reaching toward items to request objects, smiling to obtain attention, and turning in circles then gazing at the investigator to obtain attention.

During the symbol assessment conducted with Mario, he scored below 70% during the task of matching identical objects. He received a teaching session and subsequently scored 90% with the object–object and photograph–object matching tasks.

Description of Participant 2: Cody

Cody was a 4-year, 0-month-old Caucasian male with a medical diagnosis of autism. In his classroom and school setting, his educational staff used both verbal and physical assistance to help him locate scheduled activities. He used limited speech for functional communication, although he was observed to use echolalia of many single words and some two-word combinations out of context.

Results of the Communication Matrix (Rowland, 2004) indicated that Cody’s current communication abilities crossed several levels, but overall, he functioned as an intentional communicator. He used presymbolic conventional communication and early symbolic communication with emerging skills in the areas of concrete and abstract symbols. During the symbol assessment, Cody matched identical objects and matched objects to corresponding photographs with 90% accuracy.

Description of Participant 3: Sage

Sage was a 4-year, 5-month-old Ethiopian male with a medical diagnosis of autism. He exhibited particularly challenging behavior during transitions in the classroom and around the school. Typically, he was not compliant with instructions to move to a new task or location and one or two educational staff members physically assisted him during transitions.

Sage’s self-stimulatory behaviors included vocalizations, arm-flapping, and jumping. His most persistent self-stimulatory behavior was tapping objects together or against surfaces. He was nonverbal and did not imitate speech or sounds. Results using the Communication Matrix (Rowland, 2004) suggest that he was an early intentional communicator, and used unconventional behavior to indicate his wants and needs and to protest. In the symbol assessment, he matched identical objects and matched objects to corresponding photographs with 80% accuracy.

Settings

A familiar space in each school setting was selected that had a large enough area to conduct transitions across a constant distance of about 15 feet between the opening rug area and tables that were an appropriate height for the children to stand and play with the toys. With the exception of the classroom generalization phase, all sessions occurred in the same predetermined location throughout the course of the study.

Materials

The independent variable for all sessions was the type of photographic symbol presented to the child with ASD. Two types of photographic symbols of contrasting contextual complexity were used to display specific play activities on an iPad.

High-context photographs

A high-context, personally relevant photograph was taken prior to beginning each session in which the high-context condition was randomly assigned. The single photograph contained the scheduled play materials, including an essential toy structure for each activity and its accompanying parts. The high-context photographs also depicted action by the peer with the scheduled toy materials. For example, the peer peeked through the hut of the zoo set. The child with ASD was not present for this photograph, setup time.

No-context photographs

Each no-context, personally relevant photograph depicted the essential toy for each activity. The no-context photograph removed context, background context, location, partners, and active engagement elements that were depicted on the high-context counterpart in the other treatment condition. For example, the toy zoo was photographed against a black background.

iPad2

An Apple Inc. iPad2® was used during this study to display the single photographs during each condition. During both treatment conditions, the no voice-output feature was used and screen was locked to prevent the participant from rotating the symbol display. The display for the photographs in both conditions filled the entire screen of the iPad2, six-by-eight inches in size. Three identical easels for the iPad2 were placed on the table at each toy play location. A separate iPad2 was used to record each session.

Timers

The investigator wore two audible countdown timers, one to count the transition up to 3 min and the second set to count down the 4 min for toy play.

Play materials

Play materials consisted of toys that were interactive, novel, and had one essential element and several small pieces that accompanied the item. The eight toy sets selected each had 14 pieces (e.g., farm, zoo, grill).

Research Design

An alternating treatments design was used in this study. The two treatments, or independent variables, were high-context personally relevant photographic symbols and no-context personally relevant photographic symbols. Participants were concurrently exposed to the two treatments in the treatment phase. Given that children with autism are a heterogeneous group, an alternating treatment design allows participants to act as their own controls, receiving both treatments, and allows for a comparison of the relative effects of two experimental conditions on a single subject (Barlow & Hayes, 1979; Kearns, 1986). One of the basic tenets of an alternating treatment design is that the participant will experience each of the rapidly alternating treatments for an equal amount of time, with the effects on behavior noted (Barlow & Hayes, 1979). This study included a baseline phase prior to the alternating treatment phase. Although an alternating treatment design does not require a baseline, by including one, experimental control is improved by permitting a comparison of no treatment and treatment conditions.

The alternating treatment phases continued for 12 sessions. The investigator then examined the data for duration of transitions using visual inspection to select the more effective treatment. The more effective treatment phase was conducted for 3 sessions to demonstrate that the results were not affected by multiple treatment interference. An intervention was considered to be more effective if (a) it consistently produced more change or more rapid change in the target behavior for the participant, (b) stable levels of performance had been established, and (c) the majority of the data points for the intervention fall outside of the range of the values of the data points of the other interventions being studied (O’Neill, McDonnell, Billingsley, & Jenson, 2011). Upon completion of three sessions using the more effective treatment, two types of generalization probes were conducted. The first used different toys in the research setting and the second was conducted in each participant’s classroom using classroom materials.

Dependent Measures

High-context and no-context photographs were used to examine the effects of context on transitions to toy play by children with ASD. During the participants’ transition to the toy activities, dependent measures of duration, frequency of prompts, and type of prompts were coded.

Duration of transition

The duration of transition was defined as the amount of time (seconds) that elapsed starting when the investigator presented the discriminative stimulus, “time to play ________” (paired with the photograph symbol on an iPad), and ending when the participant reached the location of the scheduled activity and placed the iPad on the easel. The duration of transition was limited to a maximum of 3 min.

Prompting

If the child moved in a direction away from the toy area or if the child did not move at all, the investigator prompted the participant to move from the opening rug area to the scheduled toy play area. A hierarchy of least-to-most intrusive prompting was used to support the participants’ successful transition, using three types of prompts (symbol prompts, gesture prompts, and finally, physical guidance prompts). An explicit set of procedures was delineated to ensure consistency and fidelity of treatment. The procedures used are available upon request of the first author. The frequency of prompts was the number of total prompts given to the participant during a transition to toy play. Each prompt was coded further by prompt type.

Procedures

Overview of sessions

The two treatments of photographic symbols of contrasting contextual complexity were administered in a strict alternating fashion on the same day. The two conditions and all toy sets were counterbalanced across sessions to control for a daily order effect. The locations of play activities and peer partners were randomly assigned. Similarly, the specific play behaviors of peers depicted on the high-context photographs were randomly selected.

A session consisted of (a) a 1-min, warm-up activity, (b) presentation of one of the two types of photographs, (c) child transitioning to the depicted play activity (maximum of 3 min), (d) engaging in play with toys (4 min), and (e) a brief closing activity. A 5-min break was given, and then a second session occurred using the other photograph treatment. Each session lasted approximately 8 to 10 min in duration. Each day, one of two peers were selected and paired with the participant to engage in the toy play activities.

All phases used the same procedures for each transition to play materials and toy play. Baseline phase did not use graphic symbols for transitions, alternating treatments phase used both contrasting symbol types, and the more effective treatment and generalization phase used one of the two symbol types, based on participant data.

Environmental arrangement

The arrangement of the setting was essential to ensure that participants relied on the information from the photographs to transition, and not other factors such as a predictable task, the consistent location of a task, or the presence of peers as the cue to move to a location. The position of scheduled play activities was randomly arranged around the room to various table locations, and the peer partner was asked to remain at the opening area (rug) until the participant with ASD arrived at the scheduled table location.

Preparation of play materials

Selection of the scheduled play materials for each session was counterbalanced to control for order effects. A scheduled set of play materials was randomly selected for each session, and two distracter sets of play materials were also selected for each session. The scheduled play materials were placed at the predetermined table location for the session, also randomly selected. The two distracter sets of materials were arranged at the remaining two tables.

Baseline procedures for transitions

During baseline, the child was instructed that it was time to play with a specific toy. No photographic symbols or the iPad were used during the baseline sessions. A procedural hierarchy of least-to-most prompts and consistent time delay procedures were used. These procedures were implemented with the child during transitions to the toy activities when the child moved in a direction away from the path of the activity, or when the child did not move at all. During baseline, data collected included the frequency of prompts, type of prompts, and duration of transition (limited to 3 min).

Experimental procedures for transitions

The goal was to examine differences between photographic symbols of varying contextual complexity on the ability of children with ASD to locate and move to activities promptly and independently. Procedural steps used for transitions during the experimental phase were similar to baseline procedures. A procedural hierarchy of least-to-most prompts and consistent time delay procedures were used.

Procedures for the more effective treatment phase

During the more effective treatment phase, three sessions were conducted using the more effective treatment. The same experimental procedures were used in the alternating treatment phase.

Procedures for the toy generalization phase

At the conclusion of the more effective treatment phase, three additional sessions were conducted using toys selected from each of the participants’ classrooms. These sessions were conducted in the same location as the treatment phase of the study using the same experimental procedures.

Procedures for the classroom generalization phase

A second type of generalization was conducted in the participants’ regular classrooms, for three sessions. These sessions were conducted at the same time as free-play time for children in the classroom. Activities selected were based on teacher recommendation and the availability of toys in the classroom during the classroom generalization phase.

Procedures for the classroom generalization followed the same experimental procedures as the alternating treatment phase with the following exceptions: The opening and closing activities were not conducted, and the participants began each transition sitting in the chairs, to visually scan the activities and materials in the classroom.

Data collection procedures for duration of transition and prompting

The duration of transition and prompting for each session was coded via video. Frequency counts to score the number of prompts given as well as the type of prompting (level) necessary to support the participant to reach the scheduled play activity were coded. Duration of transition data was coded from video using a stopwatch to count the amount of time that passed between the initial instruction given by the investigator and the end of the child’s transition, signaled by the investigator stating, “ok” when the child placed the iPad on the easel.

Interobserver Reliability

Interobserver reliability was assessed via videos on a random selection of 33% of all sessions for all phases of the study. An agreement for the duration of transition occurred when both observers independently recorded the duration within a range of plus or minus 2 s. A similar measure, using a range to obtain agreement of reliability, was reported by Dettmer, Simpson, Myles, and Ganz (2000), measuring latency for transitions of students with autism. Reliability for number of prompts and the types of prompts were calculated using a point-by-point agreement method, and both agreements + disagreements were counted.

Reliability data

Reliability coding was conducted independently by a trained graduate student and the investigator. Training included two 30-min instructional sessions with the investigator, a test passed at 90% or above, and completing sample coding. Overall, interobserver reliability was 100% for duration of transitions for all three participants. Only two sessions required the range of plus or minus 2 s, and all other sessions were either an exact duration agreement, or had a 1-s difference.

Overall reliability for number and types of prompts was 98%. Reliability across participants had similar results, as Mario’s overall reliability was 98%, Cody’s overall reliability was 97%, and Sage’s overall reliability was 93%.

Procedural Fidelity

To document the integrity of the experimental procedures, data were collected using procedural integrity checklists. These checklists addressed each procedural step of the intervention. The trained observer coded randomly selected videotaped sessions using 33% of sessions across all phases. Fidelity was determined by identifying the percentage of correctly implemented procedures by the investigator.

Social Validity: Peer Perceptions of Using Photographs in the Classroom

Questionnaires were used with the six peer participants to examine any perceived changes in their perceptions of the use of photographs with their friends with ASD. The six peers were given a 6-item questionnaire, prior to beginning the treatment sessions with participants, and again at the end of the study. The investigator met individually with each of the peers and asked them to rate their agreement responses using three faces depicting emotions (happy, neutral, and sad).

Data Analysis

Visual inspection of graphed results and descriptive statistics were the primary methods used to examine changes in the dependent variables. Effect sizes were separately calculated for the graphed results of two dependent measures, duration of transitions and frequency of prompts.

The percentage of all nonoverlapping data (PAND) was calculated to determine effect size of each treatment separately. PAND was used to examine how the data differed for the participants, between the alternating treatments. To calculate PAND, a minimum number of data points that overlap are removed (across treatments) to eliminate all overlapping data points. A percentage was calculated of the remaining data points that did not overlap (Parker, Hagan-Burke, & Vannest, 2007).

To analyze the percentages using PAND, a higher percentage of nonoverlap indicates the intervention had a greater impact. A score of 90% or higher means that the intervention has been highly effective for the participants. Values ranging from 70% to 90% indicate moderate effectiveness, and values from 50% to 70% indicate minimal effectiveness. Scores below 50% indicates that the intervention had little to no effect on the participants (Scruggs & Mastropieri, 1998).

Results

Effects of Photographs of Contrasting Contextual Complexity: Duration of Transitions

Duration of transitions for Mario

The duration of Mario’s transitions resulted in a relatively flat trend line during baseline (M = 100 s, range = 77-144 s). Visual inspection of the data showed a decreasing trend for duration in both high-context (M = 50 s, range = 8-131 s) and no-context interventions (M = 25 s, range = 7-78 s), showing an intervention effect with both conditions. By the sixth session of the alternating treatment intervention, the no-context intervention had become less variable and showed stable performance by Mario. Across the alternating treatments phase, Mario consistently demonstrated quicker transitions with the no-context condition, until the final paired treatment session, when the amount of time used to transition was the same for both treatments. Overall, the no-context condition was associated with shorter durations of transitions for Mario, as compared with the high-context condition. The percentage of all nonoverlapping data (PAND) for Mario’s duration of transition across treatments was 62.5%, indicating a minimal intervention effect.

The outcome from two treatments, led the investigator to select the no-context treatment for the remaining phases of the study. Mario’s performance with the no-context photographs remained stable during the more effective treatment phase and also during the toy generalization phase. During classroom generalization, Mario’s first attempt to transition to classroom toys resulted in duration that was very close to baseline conditions. Mario’s classroom generalization showed a marked decreasing trend, with the final two sessions at the same level as his stable performance in the alternating treatments phase. Figure 1 graphically displays the data for Mario’s transitions across all phases of the study. Table 2 displays means and range for duration of transition time, in seconds, across all treatment phases for Mario.

Figure 1.

Duration of transition in seconds per session for Mario.

Table 2.

Means and Ranges of Duration in Seconds and Number of Prompts During Transitions.

			Alternating treatments
	Baseline		High-context		No-context		More effective treatment		Toy generalization		Classroom generalization
Participant dependent measures	M	Range	M	Range	M	Range	M	Range	M	Range	M	Range
Mario
Duration of transitions (s)	100	77–144	50	8–131	25	7–78	7	6–9	8	7–11	28	7–70
Number of prompts	8	6–12	4	0–12	2	0–6	0	0	0	0	2	0–6
Cody
Duration of transitions (s)	65	51–78	27	8–107	9	6–12	7	6–8	11	8–13	13	9–23
Number of prompts	5	3–6	2	0–11	0	0	0	0	0	0	.5	0–2
Sage
Duration of transitions (s)	180	180^a	20	4–46	44	7–143	8	4–10	11	5–16	14	5–20
Number of prompts	14	13–15	1	0–3	4	0–11	.3	0–1	.6	0–1	1	0–2

Sage used the maximum allowed of 180 seconds in baseline.

Duration of transitions for Cody

The mean of Cody’s baseline performance was 65 s (range = 51-78 s). On the first treatment session using the high-context photograph, Cody took 107 s to transition, well above baseline. By the second high-context treatment session, Cody’s duration of transitions dropped to baseline level (73 s), and then below baseline levels by the third high-context session, and his behavior remained stable across the entire phase of the alternating treatment condition (M = 9 s, range = 6-12 s). Visual inspection and descriptive data indicated there was a decreasing trend in duration of transitions and Cody’s duration of transitions during the high-context photograph condition (M = 27 s, range = 8-107 s) were increasingly quicker as sessions passed. By Session 7, the data indicate his behavior also became stable and at the same level as the no-context condition by the final three sessions of the treatment phase.

Overall, the no-context condition was associated with shorter durations of transitions for Cody, and remained stable, as compared with the high-context condition. The outcome from this data led the investigator to select the no-context treatment for the remaining phases of the study. The PAND across alternating treatments for Cody’s duration of transitions was 75%, indicating a moderate intervention effect. The duration data remained stable during the more effective phase (M = 7). The toy generalization phase was just slightly higher, overall, but had a slight decreasing trend (M = 11, range = 8-13). The first session of the classroom generalization phase was slightly higher (23 s), so an additional session was added to determine whether performance would become stable, which it did for the final three data points in the classroom. Figure 2 provides a graphic illustration of duration data obtained for Cody during transitions across all phases of the research. Table 2 displays means and duration of transition time, in seconds, across all treatment phases for Cody.

Figure 2.

Duration of transition in seconds per session for Cody.

Duration of transitions for Sage

During baseline conditions when no photograph symbols were used with Sage, he used the maximum amount of time of 3 min to transition to play activities and was then led by the researcher to the toys. Sage located play activities using both types of photograph symbols during the alternating treatment phase. Although Sage’s performance was variable, visual inspection and descriptive data showed a decreasing trend in both conditions, indicating an intervention effect with both treatments. The duration of transitions for the high-context condition resulted in a decreasing trend that, despite its variability, was below the no-context trend. During the alternating treatment sessions, Sage’s performance during the high-context condition (M = 20 s, range = 4-46 s) was less variable than with no-context photographs (M = 44 s, range = 7-143 s). The outcome from this data led to the selection of the high-context treatment for the remaining phases of the study with Sage. The PAND across treatments for Sage was 75%, indicating a moderate intervention effect. Sage’s performance was maintained during the more effective treatment phase using high-context photographs (M = 8 s, range = 4-10 s), and rose just slightly during the toy generalization, although the data showed a clear descending trend in duration of transitions during the toy generalization phase (M = 11 s, range = 5-16 s). The duration for classroom generalization phase (M = 14 s, range = 5-20 s) was just slightly higher than the more effective and toy generalization phases but was within the same range of data that occurred during the high-context treatment phase. Figure 3 presents graphic data for the duration of transitions for Sage. Table 2 displays means and duration of transition time, in seconds, across all treatment phases for Sage.

Figure 3.

Duration of transition in seconds per session for Sage.

Effects of Photographs of Contrasting Contextual Complexity: Number of Prompts

This study compared the impact of two types of photograph symbols on the independent ability of participants with ASD to transition to play activities. Prompting support was given to the participants each time they did not move toward the scheduled play activity. The number of prompts delivered to participants was counted for each session. Table 2 contains the mean and range of number of prompts for each participant in baseline, treatment, and generalization phases.

Number of prompts to support transitions for Mario

The data for the number of prompts needed to support Mario’s transitions closely paralleled his data for duration of transitions. The number of prompts needed during baseline ranged from 6 to 12 (M = 8), with a range of prompts during the no-context condition ranging from 0 to 6 (M = 2) and the range for high-context being 0 to 12 (M = 4). Visual inspection and descriptive data indicated that there was a decreasing trend in the number of prompts required to transition for both treatments, indicating both treatments were effective. The PAND for prompts required for Mario was 62.5%, indicating a minimal intervention effect between treatments.

The number of prompts needed during the no-context condition reached zero by the sixth session and remained stable at zero, for the remainder of treatment sessions, for the more effective treatment phase and the toy generalization phase. Six prompts were required during the first classroom generalization transition, a level the same as two transitions during the baseline phase. Mirroring the duration data, the number of prompts needed during classroom generalization dropped to zero for the final two transitions. Figure 4 displays data for the prompts across treatments for Mario.

Figure 4.

Number of prompts given to Mario per session.

Number of prompts to support transitions for Cody

The baseline data for the number of prompts used with Cody (M = 5, range = 3-6) has a similar data path as the duration of transitions graph. During the first treatment session using a high-context photograph, there was an initial spike in the number of prompts that was required for Cody to successfully transition to the play activity. Cody required 11 prompts in the first session, as compared with the highest number during baseline of 6 prompts. The number of prompts he required during the second high-context session dropped to the baseline range. During the high-context treatment condition (M = 2, range = 0-11), Cody’s performance required prompts until the fifth session, after which time no prompts were needed. Visual inspection and descriptive data indicated that there was a decreasing trend in the number of prompts needed to transition successfully during the high-context condition, indicating an intervention effect.

The no-context condition also had an intervention effect, with the number of prompts per session dropping to zero after baseline and remaining stable at this level. The PAND for Cody was 62.5%, indicating a minimal intervention effect across treatments. During treatment sessions, no prompts were needed with Cody during the no-context condition, demonstrating an immediate difference in the treatment conditions. Cody’s performance remained stable, as he moved directly to the scheduled activities depicted on the photographs, and did not require prompts during the more effective phase or toy generalization phase. The first session in the classroom generalization required two prompts, but none were needed after that session for the remaining classroom generalization sessions. Figure 5 displays the number of prompts used to support Cody’s transition to play activities.

Figure 5.

Number of prompts given to Cody per session.

Number of prompts to support transitions for Sage

Sage required several prompts during transitions to activities during baseline conditions (M = 14). Visual inspection and descriptive data indicated that there was a decreasing trend in both high-context and no-context conditions for the use of prompts. The high-context condition during the treatment phase resulted in fewer prompts needed for transitions, as compared with the no-context condition. The PAND for Sage was 75%, indicating a moderate intervention effect across treatments. During the more effective treatment phase, the toy generalization phase, and the classroom generalization phase, Sage required two or less prompts during each session. Figure 6 illustrates the number of prompts used to support Sage’s transitions to play activities.

Figure 6.

Number of prompts given to Sage per session.

Effects of Photographs of Contrasting Contextual Complexity: Type of Prompts Used to Support Transitions

This study examined the type of prompts required by participants with ASD to move to play activities. Prompts were used with participants only when needed, and the hierarchy of prompting, or type of prompt presented, began with the least intrusive and became increasingly intrusive. The type of prompt reported is the final prompt needed in the sequence of the hierarchy.

Type of prompts used to support transitions for Mario

A similar prompting hierarchy was used during the baseline phase as was during the treatment phase, with the exception of the use of photograph symbols as prompts. During baseline sessions, Mario needed eight total prompts, and when the hierarchy was used, Mario required the physical guidance support for all transitions.

During the alternating treatments phase of the study, 71% of the prompts were used during the high-context condition, as compared with 29% of total prompts used during the no-context condition. Within the high-context condition, 88% of the prompts required physical guidance, 0% of the prompts to which Mario responded were gesture prompts, and 12% of the prompts were the symbol prompts of the high-context photograph. Within the no-context condition, similar results were found, with the participant responding favorably to 14% of the symbol prompts.

Type of prompts used to support transitions for Cody

Across the three baseline sessions, Cody required five prompts, total. Four of these prompts were physical guidance (80%) and a single prompt to which he responded favorably was a gesture prompt (20%). During baseline, Cody required more intrusive support from the investigator to have successful transitions than during either of the alternating treatment conditions. When Cody was given the high-context photographs, he required physical guidance prompts until the fourth session, and responded once to a gesture prompt. After the fourth session, he responded to photograph symbols in both conditions and no additional prompting was necessary.

Type of prompts used to support transitions for Sage

Sage required 36 prompts during the baseline phase, with 39% of these prompts requiring physical guidance to reach the play location. The remaining prompts used during baseline were evenly split between expectant delay prompts and gesture prompts.

During alternating treatment conditions, 55 total prompts were used with Sage. Of these, 73% were presented in the no-context condition. In the high-context condition where 27% of the total prompts were used, Sage responded favorably to more than half of these prompts (53%) that were delivered using the photograph symbol as the prompt. Over the course of the treatment phase, he required less physical prompting, and was able to respond favorably to both symbol and gesture prompts more frequently, when prompts were needed. In the no-context condition, when Sage needed prompts (9 out of 12 sessions), it was necessary to provide physical guidance to Sage during 8 of the 9 sessions.

Treatment Fidelity

The investigator conducted all of the sessions for each participant. A trained graduate student reviewed the accuracy of implementation of the 10 procedural steps for transitions to the play activity. When overall procedures across the study phases were analyzed, there was high fidelity of implementation, at 96%, with a range of 92% to 100%.

Social Validity: Peer Perception of Photographs

Overall scores of the Social Validity Rating Scale had a possible range of 0 to 30 for each participants’ total score. All six peer participants completed the Peer Rating Scale prior to the initiation of the study and again during the generalization phase of the study. Prior to the study, peers responded neutral or negatively to five of the six questions, but half of the children (three) indicated that they would like to be in pictures used with their friend (with ASD). Individual scores prior to the study ranged from 0 to 8, with a mean of 4.5.

At the conclusion of the study, peer participants responded to the survey with more positive ratings, indicating they agreed their friend benefitted from the use of photographs, with scores ranging from 21 to 30, with a mean of 27. At the end of the study, all six participants indicated they would like to be in the pictures used with their friend (with ASD). All six peers also agreed that pictures helped their friend understand how to get to the planned play activity.

Discussion

Key findings will be discussed relative to research questions and limitations of the study, and interpretations of results will be presented with suggestions for future research.

Effects of Photographs on Duration of Transitions

The impact of two types of photographs of contrasting contextual complexity were examined in relation to the amount of time children with ASD needed to locate play activities. All three participants moved to play activities more quickly using both types of photograph symbols, as compared with their duration of transitions during baseline. These results are consistent with existing literature that demonstrates visual supports lead to positive outcomes related to independent transitioning (Carson, Gast, & Ayres, 2008), and a decrease in latency between instructions and when children begin tasks when visual supports are used (Dettmer et al., 2000).

Two participants transitioned more quickly to play activities when given the no-context photographs, as compared with the high-context photographs. In contrast, the third participant, Sage, transitioned more quickly using high-context photographs. Differences between participant characteristics could account for the type of photograph affecting the duration of transitions to play activities. Assessment scores of Sage, the participant who located play activities faster using high-context photographs, reflect a child who has communication and adaptive behavior skills that are less well-developed than the other two participants. His performance matching objects to photographs was also lower than scores of the other two participants. Additional research is needed to explore matching participant characteristics and examine the impact of context for children having more significant challenges.

It is possible that individuals with ASD, who experience more cognitive challenges, may benefit from contextual photographs, but there is a paucity of research specific to individuals with ASD. Some leading researchers postulate that contextual photographs, or VSDs, reduce cognitive demands, and offer immediate success using the photographs (Blackstone, 2004). Shane and Weiss-Kapp (2008) suggest that whole scenes are likely to provide greater meaning of schematic or experiential content to persons with ASD including some who may not comprehend the semantic meaning behind images of isolated nouns.

Wilkinson and Jagaraoo (2004) compared high-context photographs with iconic symbols, and their research findings demonstrated that people process high-context photographs more rapidly than iconic symbols. The authors suggest the increased processing speed of high-context photographs could reflect a decreased cognitive workload. Individuals with ASD who are more cognitively impaired, similar to Sage, may be more successful, and respond more quickly, using contextual photographs that decrease their cognitive workload.

Effects of Photographs on the Number of Prompts Provided to Participants

This study also examined the impact of two types of photograph symbols on the independence of participants with ASD to transition to play activities, measured by the number of prompts required to reach the scheduled location. All three participants moved to play activities with less prompts using both types of photograph symbols, as compared with the number of prompts required during baseline. This finding is supported in the literature, based on positive outcomes for individuals with ASD using visual supports during transitions. These outcomes have been measured by a reduction in the use of verbal, gestural, and physical prompts (Dettmer et al., 2000; Hall, McClannahan, & Krantz, 1995; Massey & Wheeler, 2000) given to individuals with ASD during transitions.

The results in this study for the number of prompts required by participants paralleled the findings for the duration of transitions. Two participants performed better during the no-context condition, Mario requiring less prompts than in the high-context condition, and Cody did not need any prompts to support his transitions. Similar to the results for duration of transitions, Sage is in contrast to the other two participants, and required less prompts using high-context photographs, compared with no-context photographs. Cody and Mario had minimal effect size differences between treatments; however, Sage’s data showed a moderate effect size. For Sage, the type of photograph used was more important for his performance than the other two participants who had higher skills, based on assessments. Future investigations should be conducted to examine the features in photographs that provide the most salient contextual information to individual children with ASD.

The investigator conducted three baseline sessions with participants prior to paired sessions of alternating treatments. Immediately following baseline, two of the three participants, Sage and Cody, exhibited an immediate reduction in the number of prompts required to transition to the scheduled location of the play activity. This immediate difference was more prominent with the condition that was more effective, overall, across the alternating treatment phase. The more effective treatment for Sage was high-context, and was no-context for Cody. Mario’s reduced need for prompts was less notable, although his declining trend begins well below his baseline trend for the no-context condition (more effective). In this study, the only procedural differences between baseline and treatment sessions were the introduction of photograph symbols on the iPad2 and the use of the photograph as an initial prompt.

Existing research demonstrates the effectiveness of visual supports to improve on-task behaviors (Spriggs, Gast, & Ayres, 2007), self-management (Newman et al., 1995), and independent engagement in tasks (MacDuff, Krantz, & McClannahan, 1993; McClannahan & Krantz, 1999). These studies used either an instructional component or reinforcement to teach schedule-following behavior using visual supports. The present study did not use direct instruction, modeling, or reinforcement with participants, but the provision of specific types of photograph symbols may have provided a visual support as augmented input. The type of photograph, alone, may have contributed to the more independent performance of participants in this study.

Effects of Photographs on the Type of Prompts Required by Participants

The type of prompts required by participants with ASD to move to play activities was examined. When participants required prompts, these were delivered using a least-to-most intrusive hierarchy. Cody did not require any prompts in the no-context condition; therefore, no data for prompt type are available for him in that condition.

The type of prompts required became less intrusive for two of the participants, Mario and Sage, in the high-context condition. Cody needed intrusive prompts for a few sessions in the high-context condition and then did not need any prompts. Using the high-context photographs, both Sage and Mario demonstrated a more gradual reduction in prompt intrusiveness, and both also responded to the high-context photograph as a prompt. Their success in locating the scheduled activity when prompted to orient toward and gaze at the high-context photograph indicates this type of symbol was meaningful to these participants, who used the contextual information to transition successfully.

Sage had the most notable difference in the type of prompts used to support his transitions. Using the no-context photographs, it was necessary to provide full physical guidance to Sage for 8 of the 12 sessions, and he did not successfully respond to the no-context symbol as a cue to locate the scheduled activity in any of the treatment sessions. In contrast, using the high-context photographs, Sage required physical guidance for only 2 of the first 4 sessions, and responded positively to the photograph symbol as a cue to locate the scheduled activity. This marked difference between the two symbol types may have demonstrated he used the information in the high-context photographs. Based on his performance using the high-context symbol, the high-context symbol acted as a cue to help Sage be successful.

Limitations

The procedures used to begin the transition to the toy play area included the verbal instruction, “time to play _____.” The play structure was always named with specificity, such as “time to play farm” or “time to play zoo.” It is possible that the participants learned the spoken vocabulary about of the play structure and transitioned to the toy set based on the verbal information, not due to the symbol cue. However, because baseline was included in the design of the study, and no graphic symbols could be given, it was necessary to use a verbal instruction, which then was continued for all sessions. It is difficult to predict whether it was the symbol alone that impacted the participants’ behavior or the language used by the investigator.

Any single toy set could have been displayed as a high-context or no-context photograph. It is conceivable that a participant may have learned to discriminate a specific toy via one of the conditions and remembered the toy in the other condition. All toys were randomly selected and counterbalanced across conditions, to reduce this influence.

After baseline, two alternating treatments were compared by delivering an equal number of sessions in both conditions, and the more effective condition was selected for the next phase. The addition of baseline was to enhance the traditional alternating treatment design. Based on previous research, we acknowledge that children with ASD may be positively influenced by the use of any augmented input. This may have accounted for the precipitous decrease from baseline to the alternating treatment phases. Future replication research should be conducted with focus on baseline as a treatment condition.

Directions for Future Research

The portability, ease, accessibility and widespread use of iPads make the device a logical choice to support children with ASD. Future research that focuses on contextual photograph symbols could use an application that has a dynamic display to decrease the static nature of the photo display. For example, one photo might depict the destination in the classroom, and the next photo could depict a high-context photograph of what to do at the destination. Mario’s outcomes provide a direction for future research. During the final three sessions, Mario connected with the idea of imitating the photograph displayed on the iPad2. Mario carried the iPad2 to the table of toys, placed it in the easel, performed the high-context behavior depicted in the display, and then stood and watched his peer. Future research should explore if children with ASD would stay more actively engaged if the contextual photo display was dynamic, or if there was a series of contextual photos.

Future research should investigate the use of high-context photographs in the classroom setting. Toys used in this study were selected for their interactive nature and sensory feedback. Some materials in classrooms may not be as interactive as the ones selected for this study. Future research that examines the use of contextual photographs using standard classroom materials may be valuable. Moreover, research about contextual photographs should be done in the classroom context, and research about transition to activities should occur during typical junctures of classrooms.

Additional research is needed to explore how the contextual complexity of photographs differs across children who have different learning needs. Specifically, more challenged learners may benefit from high-context photographs to transition to tasks independently and engage in social interactions with peers. It may be beneficial to include children with more challenging needs in research that examines opportunities to use photograph symbol supports.

Footnotes

Acknowledgements

The authors would like to acknowledge the participation of Lincoln Public School students, teachers, and administrators and Mary Ells and Mary Phillips, in particular, for making this study possible.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

References

American Psychiatric Association. (2013). Diagnostic and statistical manual of mental disorders (5th ed.). Washington, DC: Author.

Barlow

D. H.

Hayes

S. C.

(1979). Alternating treatments design: One strategy for comparing the effects of two treatments in a single subject. Journal of Applied Behavior Analysis, 12, 199–210.

Beukelman

D. R.

Mirenda

(2013). Augmentative and alternative communication: Supporting children and adults with complex communication needs (4th ed.). Baltimore, MD: Paul H. Brookes.

Blackstone

(2004). Visual scene displays. Augmentative Communication News, 16, 1–8.

Carson

K. D.

Gast

D. L.

Ayres

K. M.

(2008). Effects of a photo activity schedule book on independent task changes by students with intellectual disabilities in community and school job sites. European Journal of Special Needs Education, 23, 269–279.

Dettmer

Simpson

R. L.

Myles

B. S.

Ganz

J. B.

(2000). The use of visual supports to facilitate transitions of students with autism. Focus on Autism and Other Developmental Disabilities, 15, 163–169.

Hall

L. J.

McClannahan

L. E.

Krantz

P. J.

(1995). Promoting independence in integrated classrooms by teaching aides to use activity schedules and decreased prompts. Education and Training in Mental Retardation and Developmental Disabilities, 30, 208–217.

Kearns

K. P.

(1986). Flexibility of single-subject experimental designs. Part II: Design selection and arrangement of experimental phases. Journal of Speech and Hearing Disorders, 51, 204–214.

MacDuff

G. S.

Krantz

P. J.

McClannahan

L. E.

(1993). Teaching children with autism to use photographic activity schedules: Maintenance and generalization of complex response chains. Journal of Applied Behavior Analysis, 26, 89–97.

10.

Massey

N. G.

Wheeler

(2000). Acquisition and generalization of activity schedules and their effects on task engagement in a young girl with autism in an inclusive preschool classroom. Education and Training in Mental Retardation and Developmental Disabilities, 35, 326–335.

11.

McClannahan

L. E.

Krantz

P. J.

(1999). Activity schedules for children with autism: Teaching independent behavior. Bethesda, MD: Woodbine House.

12.

Mirenda

Brown

K. E.

(2009). A picture is worth a thousand words. In Mirenda

Iacono

(Eds.), Autism spectrum disorders and AAC (pp. 303–332). Baltimore, MD: Paul H. Brookes.

13.

Mirenda

Locke

P. A.

(1989). A comparison of symbol transparency in nonspeaking persons with intellectual disabilities. Journal of Speech of Hearing Disorders, 54, 131–140.

14.

Newman

Buffington

D. M.

O’Grady

M. A.

McDonald

M. E.

Poulson

C. L.

Hemmes

N. S.

(1995). Self-management of schedule-following in three teenagers with autism. Behavior Disorders, 20, 190–196.

15.

O’Neill

R. E.

McDonnell

J. J.

Billingsley

F. F.

Jenson

W. R.

(2011). Single case research designs in educational and community settings. Boston, MA: Pearson.

16.

Parker

R. I.

Hagan-Burke

Vannest

(2007). Percent of all non-overlapping data (PAND): An alternative to PND. The Journal of Special Education, 40, 194–204.

17.

Prizant

(1983). Language and communicative behavior in autism: Toward and understanding of the “whole” of it. Journal of Speech and Hearing Disorders, 46, 241–249.

18.

Rowland

(2004). Communication matrix. Unpublished manuscript, Health & Science University, Portland, OR.

19.

Schopler

Mesibov

G. B.

Hearsey

(1995). Structured teaching in the TEACCH system. In Shopler

Mesibov

G. B.

(Eds.), Learning and cognition in autism (pp. 243–267). New York, NY: Plenum.

20.

Scruggs

T. E.

Mastropieri

M. A.

(1998). Summarizing single-subject research: Issues and applications. Behavioral Modification, 22, 221–242.

21.

Shane

H. C.

(2006). Using visual scene displays to improve communication and communication instruction in persons with autism spectrum disorders. Perspectives on Augmentative and Alternative Communication, 15, 7–13.

22.

Shane

H. C.

Weiss-Kapp

(2008). Visual language in autism. San Diego, CA: Plural.

23.

Sparrow

S. S.

Cicchetti

D. V.

Balla

D. A.

(2005). Vineland Adaptive Behavior Scales: Second Edition (Vineland II), Survey Interview Form/Caregiver Rating Form. Livonia, MN: Pearson Assessments.

24.

Spriggs

Gast

Ayres

(2007). Using picture activity schedule books to increase on-schedule and on-task behaviors. Education and Training in Developmental Disabilities, 42, 209–223.

25.

Weissling

Beukelman

(2006). Visual scene displays: Low tech options. Perspectives in AAC, 4, 15–17.

26.

Wilkinson

K. M.

Jagaraoo

(2004). Contributions of principles of visual cognitive science to AAC system display design. Augmentative and Alternative Communication, 20, 123–136.

27.

Wood

L. A.

Lasker

Siegel-Causey

Beukelman

D. R.

Ball

(1998). Input framework for augmentative and alternative communication. Augmentative and Alternative Communication, 14, 261–268.